The epidemiology of adolescent AOD use and our understanding of the trajectories of use are, to a great extent, reliant upon recall and self-report. This study will use the NLSY child data from 1988-2000 to examine report consistency for adolescent AOD use across seven waves of data spanning 13 years. We propose to utilize both the Interference Theory and Length of Time Theory of memory and recall to guide our multi-wave analyses to better understand changes in reports of lifetime alcohol and drug use. Interference theory posits that the number of similar or related events that occur to an individual impacts their ability to recall accurately prior events. Further, we will examine the changes in reported ages of onset for alcohol and other drugs across time using the Length of Time theory which includes recall decay and telescoping. We will test for telescoping error that is, reporting an age of onset younger or older than the originally reported age of onset, in order to identify patterns and predictors of changes in ages of onset for each substance reported. In order to accomplish these overarching goals we propose the following specific aims: 1) To examine the changes in the reports of lifetime use and ages of onset for the substances queried wave by wave. 2). To identify predictors of differences in report stability of: a) ever use and b) ages at onset across sex, SES, ethnic groups, number of other drugs reported, number of years since age of onset, and frequency of use. 3). To identify if changes in report stability of use are impacted by time since the identified event is first reported based upon the Length of Time theory. 4). Test the Interference theory of memory, and specifically the proactive and retroactive inhibition theories to examine stability of recall. One Exploratory Aim: Identify patterns of substance use reports over time and examine if there are typologies of recanting that can be revealed in the data from waves 4-7 for alcohol and 5-7 for cigarette and marijuana use. Seven publications have been outlined that will result from these aims. The publications will result from wave-by-wave analyses;longitudinal, multivariate analyses and latant class analyses. Public Health Relevance: In order to accurately target appropriate age groups for AOD prevention efforts we must know when youth truly begin use since adult recall of use is at risk for multiple biases. This would be one of the only projects to follow a cohort of children through young adulthood in order to examine changes in self-reported substance use and ages of onset. Furthermore, this will be the only research to be driven by memory theories. Having the information from this study will help the field to more accurately estimate the prevalence of substance use among children and young adults but it would also inform the field as to when we should offer children prevention programs.